/diy/ - Do It Yourself

>>I have junk, what do?
What you are referring to as junk is one day's potential treasure.

>>1731558
that's meant to refer to the people who come in here with some piece of consumer electronics and want something "cool" to do with it

>>1731520
this thread's digits brought to you in part by the HMCAD1520 1Gsps ADC

https://www.st.com/resource/en/data_brief/steval-mki193v1.pdf

I have this thing and I'm trying to hook it up to an STM32 Nucleo board I've got, but I have no idea how to power it. The only thing relating to power I see is I think it's saying use a 100nF capacitor? I'm very new to this so I'm not sure if I'm supposed to be able to figure out how much power to give it from that. I guess I just have to figure out how much I can give that capacitor without it exploding, and the capacitor will handle the rest? Can anyone explain how to not fry this thing?

>>1731631
welcum to the club buddy.
What you're looking at in that link is an evaluation board. What you see in that datasheet is everything that ST has already laid out for you (like the 100nF cap). What you're looking for is the technical specs for that chip (U1) which can be found on the schematic page.
>the solution
Google the part number for that chip (Axxx330 or whatever), and view the datasheet; what you're technically looking for is pic related (VDD)

>>1731665
Ohh, okay that makes more sense. I found the datasheet you posted a pic of. So, 3.3v should be good, right? I think the Nucleo board has a 3.3v pin.

>>1731671
Yes

this is a classic hex oscillator. You can control a neat sounding effect by varying the resistance of the right resistor. http://tiny url com/wvub22a

my thing is this: Obviously it's well and dandy that I can control the resistance by hand. I want to modulate the resistance via another signal (like a sine or saw LFO). I have no fucking clue how. I *thought* that was exactly what a transistor was for, but I tried substituting an npn transistor in place of the resistor, and putting a saw sweep on the base, but it did nothing at all.

What is the standard way of modulating resistance (or current flow I guess) via another signal?
And maybe, what is the abstraction above that? What is the general solution to trying to modulate carrier X with signal Y when it comes to electronics?

>>1731715
JFETs act more like resistors than BJTs but, you really want a voltage controlled oscillator.

>>1731715
JFET or maybe an LDR+LED combo like some effects pedals use. But I'd recommend keeping the resistor constant and instead using a varactor circuit. Simpler and cheaper than using a JFET.

>>1731724
I understand that the VCO's gonna provide the modulating signal, but where does it actually modulate the resistance?

>>1731715
Basically what other guy said can work, but it would be best if you just bought an analog multiplier (which uses the gilbert cell you see here). perhaps a combo would be best
http://tinyurl com ua5e5d9

>audio output
oh man, spice should have this

>>1731734
I would pay a concerning amount of money for a standalone, high-performance souped up version of falstad. Spice-based editors are just terrible for intuition and simple stuff involving audio.

>>1731739
Have you considered that you're just a brainlet?

>>1731715
what, this?
https://en.wikipedia.org/wiki/Oscillator_sync

>>1731728
unfortunately, this is the most correct answer

>>1731739
1. build it with electron
2. throw hardware at it

>>1731746
>calls other people brainlets
>while namefagging

Good MOOCs, or books that will tech me enough about PCB design so I can freelance for extra cash?

What is pic

>>1731815
>I want to buy experience
you want experience? well, experience costs. and right here's where you start paying

>>1731851
Bro what

>>1731824
is smol birb with 0201 passives

>>1731854
git gud

>>1731856
im try

>>1731715
> What is the standard way of modulating resistance (or current flow I guess) via another signal?
The most common is a FET, which acts much like a voltage-controlled resistor when not saturated. However, a FET only has 3 terminals, so unless the source is at a (relatively-) fixed voltage you also need an adder (Vg=Vs+Vctrl => Vgs=Vg-Vs=Vctrl).

Light-dependent resistors are a common alternative if you need 4 terminals or isolation and you don't need a fast response or sale in the EU (cheap LDRs are cadmium-based, which is barred by RoHS for non-essential applications and "I can't figure out how to drive a JFET or OTA" isn't considered a valid reason for an exemption).

> What is the general solution to trying to modulate carrier X with signal Y when it comes to electronics?
It depends upon the details. At audio frequencies, the sane solution for any non-trivial signal processing is DSP. The only reason for choosing an analog approach is if your market consists of hipsters who want something to go with their valve (tube) amp and vinyl records.

>>1731728
Oh also operational transconductance amplifiers or even mixers might be able to work here. Not that I'd recommend either of them, but they could work.

I have a relay which has a coil that needs about 100ms pulse of 12V DC in order to trigger.
But, my power source is only 3V, what is the easiest way to trigger this relay? Higher voltage power source is not possible (i am using a 2xaa battery)
So far the only solution i can think of is to add a boost converter hat on top of my thing, but that is such a huge fucking overkill just to trigger one relay

>>1731876
>need more voltage
Change power supply
>can't change power supply
Boost converter
>boost converter would be overkill for just one relay!
Change relay or add more 12VDC components to justify a boost converter

>>1731884
is there no other way to raise voltage except a boost converter? how about some shit with coils?

>>1731886
>how about some shit with coils?
like a boost converter?
or do you mean a transformer?

>>1731715
People sugesting Jfets are correct but they are hard to find these days. Take a LED and a LDR, put them face to face inside a heatshrink tube, or any dark tube, now you have a resistor that can be controlled by the led.

>>1731899
and you asked about what is used to multiply two signals. Well.. DSP teacher make it sound very easy by just placing a mutiplier block. But usually that is a variant of something called gilbert cell.

Any idea where I can source very fine quartz filaments, preferably coated with silver? I've been reading about Einthoven's galvanometer lately.

>Einthoven's invention consisted of a silver-coated quartz filament of a few centimeters and negligible mass that conducted the electrical currents from the heart. This filament was acted upon by powerful electromagnets positioned either side of it, which caused sideways displacement of the filament in proportion to the current carried due to the electromagnetic field. The movement in the filament was heavily magnified and projected through a thin slot onto a moving photographic plate.

Sounds like a fun experiment, though I certainly wouldn't need to use a photographic plate. I have no clue where to source fine quartz filaments though.

I'm pretty new to the whole /ohm/ stuff, and I want to power multiple different chips that need different voltages DC. I'd want to have one plug only for the system. The input would be the wall plug, so (since I'm in europe) 230 V ± 23 V , at 50 Hz.

I thought the best idea would be to use transformers, then add full bridge rectifiers behind, and capacitors after these to reduce voltage spikes. Does that sound reasonable or dumb?

>>1731913
You could 100% just take a computer power supply and convert it for something like this. I'm guessing you need +12, +5, and maybe +3.3? With a computer power supply you'd also get -12, and on older supplies you'd also have -5.

Here's a video showing how to turn one into a bench power supply: https://youtu.be/z2oSFpKh_Uw

>>1731913
>Does that sound reasonable or dumb?

that's the standard way to do this sorta thing. but you dont necessarily need more than one xformer. typically, you'd use several regulator chips to derive the necessary voltages from a single xformer/rectifier/capacitor combo.

To handle static emi does grounding to wall wart ground (the same wall wart that powers the circuit) do anything or does it HAVE to be earth ground?

>>1731876
make your own boost converter
but it's better to change relay

>>1731899
>treating alibay as the horizon of the industry
every online distributor worth a fuck has a whole section of them

>>1731900
just use an OTA

>>1732018
the mains is a low impedance source no matter which terminal you connect to

>>1732064
>the mains is a low impedance source no matter which terminal you connect to
the problem is i don't want to ground to mains, i want to ground to the - terminal of the wall wart because i am too scared to ground to the actual mains earth ground

>>1732064
>industry
>a single dude doing something
Ok. But they are also hard to come by here in the third world too. You can get them sure, but for 10x more than a OTA, optocoupler or a LDR and LED.

I made a post in the previous thread, and I solved the problem. It wasn't a big one (I was just being a little stupid) but I'll leave it here.

I had a board with a buck converter as a seocndary power supply, and it was working half a second before stopping and restarting in loop. The controller is a Viper013hs. I think it was broken. It wasn't.
Well, I am ocasionally repairing boards with failed power supplies, and they are always failed buck converter.
Everytime, I connect the power pins to a power supply up to 63V. It always work, it's a pretty low voltage, but I never had problems with it. I have powered TVs with that PS.
Well, this time, it wasn't working because of my power supply. The IC was restarting because the output wasn't achieving the needed voltage for the feedback pin, and it interpreted that low voltage as a shortcircuit or overcharge.
Connecting the board to mains power, it worked just fine.

Pic related is a long exposure photo (5s) of my scope displaying the output voltage (up) and the voltage on the IC feedback pin (down).

why is the complexity of everything electronic absolutely exponential?
you start off with [basic circuit] using like maybe a transistor and a few capacitors and decide "hey I wanna add [presumably simple concept] to it" and it turns out you'll need several expensive specialty components, a microcontroller, a million auxiliary components, and several positive/negative voltage power supplies at various gradations

obviously that's hyperbole but it feels like whenever I want to actually *do* anything of note with electronics the difficulty and complexity ramps up hilariously fast

>>1732097
Look into how the same task might've been done in years past, there might be something that uses a few simple discrete components, but just doesn't provide control or accuracy to the level the IC solution would.

That said, sometimes when you look into how shit used to be done, you find them using unusual materials and setups. Take the original EKG machine. It used a fine quartz filament, plated with silver, suspended between two high-power electromagnets. The two EKG terminals across which the voltage was to be measured would be connected to this filament, which would deflect as current passed through it.

I mean, electromagnets are reasonably cheap to get, or make, and the electrodes were just pails of saline (i.e., salt water). But that fucking filament. I cannot figure out where to source fine quartz filaments, let alone silver-plated ones. I might be able to use platinum wire instead, but I believe I read the quality of the reading isn't as good.

>>1732016
Thanks for the info, got some reading up to do.

>>1731899
> People sugesting Jfets are correct but they are hard to find these days. Take a LED and a LDR
Farnell lists 120 JFETs vs 9 LDRs (none of which are RoHS compliant). JFETs start at £0.06 (unit qty), LDRs at £0.99.

Are you looking at "hobbyist" suppliers? LDRs are pretty niche items; compared to photodiodes and phototransistors, they're expensive, slow, and temperature dependent (and, in the EU, regulated to about the same level as nerve agents and nuclear weapon components). The one thing they have in their favour is that they're usable by someone whose knowledge of electronics boils down to V=IR and knowing not to hold a soldering iron by the hot part.

JFETs have far more practical uses, but you need a basic understanding of circuit theory (first-semester EE, or ham radio level) to make use of them. Unlike BJTs or MOSFETs, they aren't much use in the "how to drive a 1A load from an Arduino pin" department.

>>1732097
shit dude you only got 1 INT

>>1732158
fill me in on JFETs. they just seem like symmetric MOSFETs with lower noise.
t. wrapping up first semester of EE

for those of you who are already accomplished with electronics, when you're looking at an arbitrary circuit for the first time...
Do you analyze it in terms of trying to figure out where the current goes from the positive rail to ground, and how it interacts with each component?
Or do you just recognize the forms of common component groups and their functions, and see it as a black box kind of thing, disregarding the low level stuff?

>>1732166
Compared to MOSFETs, JFETs are typically low-power devices, with higher gain and lower noise. They don't have a body diode so they can be used as bidirectional (AC) switches.

The main functional difference is that JFETs are depletion-mode devices: they conduct when Vgs=0 with conduction reduced as Vgs becomes more negative (for n-channel, positive for p-channel).

The fact that Vgs=0 is in the linear region allows them to be used in very low voltage applications, e.g. energy-harvesting circuits which are basically a boost converter which can operate from an input of tens of mV.

>>1732080
it's *your* funeral, friendo
>don't want to
everyone laugh at his pain when he plugs in the wall wart the wrong way round
as for you, look up Y capacitors

>>1732081
maybe not so easy to *get* in all their flavors, but you can *find* them easily. they're not a commonly used part, to be fair, and there are usually better ways to do it than JFETs e.g. MOSFETs may serve as well or better, or building the amplifier yourself out of discretes because you don't need the accuracy or the shit of the $2 or more part. matched BJTs in a single SOT-363 package aren't that ridiculously expensive, where their guaranteed close matching is really needed

>>1732095
9/10 bench photo. I think that should be the next OP

>>1732097
either because I'm tripping over my own dick and should have started with a clean sheet instead of hacking the existing design, or because I'm working it the wrong way round and should have started with a different part of the concept

>>1732174
>Do you analyze it in terms of trying to figure out where the current goes from the positive rail to ground, and how it interacts with each component?
almost never. most interesting circuits have input(s) and output(s), so I start at the input(s) and work my way to outputs to get a qualitative sense of how one changes with the other, mostly disregarding small errors if I can. most useful circuits are also somewhat symmetrical. power can be taken for granted until I have a reason to care about it
>Or do you just recognize the forms of common component groups and their functions, and see it as a black box kind of thing, disregarding the low level stuff?
something like this, with the black box to be illuminated later as needed. you can get far by assembling standard cells of 2-6 transistors at the next higher level of structure, witness the semi-custom analog chip industry

>>1732177
just saying, depletion-mode MOSFETS do exist, mostly found in RF stuff

Something I only recently figured out: The reason they're called Vcc, Vee, Vss, Vdd: It follows from the practice of pluralizing a shorthand by doubling it. So, just as "page" is abbreviated "p" and "pages" is abbreviated "pp", "section" as "§" and "sections" as "§§", and "paragraph" as "¶" and "paragraphs" as "¶¶", the Vcc, Vee, Vss and Vdd lines are the voltage rails for collector*s*, emitter*s*, source*s*, and drain*s*.

>>1732158
Sorry m8, I buy my shit from locals or aliexpress. It's just that the electronic market in brazil is 99% for appliance repair and 1% overpriced "maker" stuff. Most store pages here list the type of the transistor as "black", "SOT" or some retarded shit and people get by part numbers because they are mostly repair techs changing a busted part. Here a matched pair of JFETs I used in a pedal were about 5$ and a LDR was scavanged from a garden lamp but can be found for 0.2$ or less.

>>1732158
> (and, in the EU, regulated to about the same level as nerve agents and nuclear weapon components)
I call bullshit on this. What the fuck

>>1732189
Kek sounds like Radio Shack in the US back in the 90s and 2000s. They'd have drawers in the back full of components like how hardware stores have drawers for different fasteners. It was kind of embarrassing.

>>1732191
It's specifically ones that use cadmium in the semiconductor portion, which I think are the most common and inexpensive type.

I actually had no idea they had cadmium in them. I'm gonna give the ones I find a wider berth from now on. I always hate when I find a melted down NiCd battery because of how nasty that shit is.

>>1732191
RoHS is a harsh mistress

>>1732193
I was in there all the time in the 1980s and thereabouts and they had plenty of blister-carded components of the most common varieties. as did my local brick-and-mortar independent electronics distributor with literally aisles of Jim-Pak stuff and plenty of ECG stock available from the TV repair shop adjoined to it. I started doing Digi-Key instead around 1990 or so when I got serious about the hobby

>>1732193
So how is it now?

>>1732214
I think they closed all the RadioShack stores in my state last time they went through bankruptcy.

Anything I should know about these soldering irons?

>>1732221
I'd just get one on ali express. Bigclive did a review of them awhile back and they seem perfectly useable, though the temp control won't be accurate out of the box (probably just a linear offset).

>>1732221
get a T12 chink station instead

>>1732193
Honestly I long to see the day when local brazilian suppliers actually start sorting things by function instead of part name.

brushless dc motor with liquid cooling through the stator
thoughts?

>>1732221
weller's only unique product is the magnastat. there's no reason to buy anything else from them when you can get the same shit unbranded for less money.

>>1732261
Isn't that just a standard curie-point thermal switch?

>>1732191
It's hyperbole.

Most visible-light LDRs use cadmium sulphide (CdS) or cadmium selenide (CdSe). RoHS (Restrictions on Hazardous Substances, the same regulation which prohibits lead in solder) prohibits the use of cadmium in concentrations above 100ppm for most uses (there's an exemption for CdTe solar cells). RoHS doesn't apply to batteries (there's a separate, less-restrictive directive for that), so NiCd batteries are legal (but have largely been superseded by NiMH for consumer applications).

So "same level as nerve agents and nuclear weapon components" is a gross exaggeration, but you aren't going to sell a product containing CdS LDRs in the EU. Some manufacturers of (niche) audio hardware (which sometimes uses LED+LDR as a low-noise voltage-controlled resistor) filed for an exemption but were denied.

>>1732288
kek oh shit i forgot that optos would be covered by that shit.
oh my god what do euros do? magnetically couple everything?

>>1732292
nah optos tend to use aluminium gallium arsenide which is apparently fine under RoHS

>>1732219
I don't think any Radio Shack stores were left open except maybe for "Dealer" stores. Seems like a lot of the remaining NOS parts after the closeout sales went to Hobby Town stores.

>>1732257
in the mean time you can borrow the parametric selectors of maunfacturer and distributor web sites and maybe learn a few part numbers/marking codes that you expect to use frequently. e.g. in SOT-23 or SOT-323 packages, "1A" is almost always something like a 2N3904

>>1732296
he means LDR optocouplers

>>1732292
MOSFETs, JFETs, OTAs, or good old DSP

>>1732267
yep. it's a real hoot when the mechanism gets stuck. lights up like a glowstick!

Does valve stuff go here too or is there a dedicated thread for older gear?
https://www.youtube.com/watch?v=trYTsCg7SWE

>>1732461
It's welcome here, just somewhat of a rarity besides audiofool stuff. The more obscure the part, the better.

>>1732308
> he means LDR optocouplers
Which were never really all that common. It seems to be mainly an audiophool thing. The alternatives tend to be better in every regard bar simplicity. The main thing LDRs have going for them is that you can take an existing circuit and replace a fixed resistor (or rheostat) with an LDR without even needing to understand the original circuit, let alone such arcane wizardry as how to use a FET.

I suspect the replacement probably involves wrapping a thermistor in nichrome. Or bolting a servo onto a physical pot.

>>1732480
>better
they were worse at being floating voltage-controlled resistors. VCO/VCA/VCF aside, how would you have done effects like Pic related in 1970 without them?

Has anyone tried to make a distortion pedal by having a piezo make sound travel through a mechanical element like a spring and having another piezo at the other end picking up the resulting sound?
I thought it was a neat idea.

>>1732492
I think something like that is used for echo/delay pedals, as opposed to distortion ones.
Makes me wonder if there's a mechanical way of mixing/multiplying two functions?

>>1732181
>everyone laugh at his pain when he plugs in the wall wart the wrong way round
no, you don't understand, i will be using the wall wart DC ground NOT the mains earth ground (and not the neutral wire which isn't even ground)

Im going to try and make my first pcb but i have a doubt regarding ferric chloride. How long do you guys soak the board on the ferric chloride? and its always necesary to heat the solution first?

>>1731715
easiest way without using any kind of controller is ldr.

I found this in a dumpster, any idea what it is?

I dumpster dived it because there are 10 lm317s on it, fucking jackpot

>>1732308
>in the mean time you can borrow the parametric selectors of maunfacturer and distributor web sites and maybe learn a few part numbers/marking codes that you expect to use frequently. e.g. in SOT-23 or SOT-323 packages, "1A" is almost always something like a 2N3904
Yeah, that is what I do. Pain in the ass to crosscheck to web searches tho.
If LDRs are haram, what euros use in their garden lights, timed light fixtures etc ?

>>1732492
Yes. It worked, I used two speakers for tranducers tho.
>>1732497
Yes. Search for mechanical computer on youtube. Really cool old military videos explaining fire control computers for ships.

>>1732532
>yes it worked
And old reverb pedals work exactly licke that but with a custom spring and magnet contraption.

>>1732529
An industrial PID controller

when is the best time to dumpster dive my college for electronic equipment

>>1732529
>NEC D7002C Dip-28 12-bit Binary A/d Converter
>those analog daughterboards with LM119D dual comparators
>no obvious CPU (sticker covering that one chip, it's pretty old if it's a microcontroller)
You say those are LM317 voltage regulators? They have no heat sink, so they're either for reference voltages or trying to keep each channel on a separate regulator.
Maybe it's some kind of data acquisition board?

>>1732530
>If LDRs are haram, what euros use in their garden lights, timed light fixtures etc ?
I get junk garden lights all the time from by-the-pound thrift stores, and take them apart partly for fun, and partly to bodge up solar panels for fun. A few years ago the CdS versions all disappeared; NiCd batteries got replaced by NiMH soon after that. Now they use solar cell voltage to know when to turn the light on, the chinks are probably happy to use one less part.

I did notice that some of them used a 600MAh AA cell, holy crap that's weak. A good NiMH cell now is over 2000.

>>1732541
When labs throw their shit out. I got 10 20$ IGBTs that way out of stupid post doc converter projects. and a lot of magnetic cores

>>1732492
...A spring reverb?

Before CCDs, how was live TV implemented? How did space program live broadcasts were made?
>implying the moon landing is real

>>1732598
I'm asking more of how the fuck was image converted to electrical signals. The broadcast was probably through SSTV right?

>>1732504
that's just 0V and not a very low impedance one at that. it would be mostly useless for EMI earthing. you will just make a bigger antenna>>1732519

>>1732519
until it's done. you have to tune the process in your own lab which depends on the boards themselves, copper thickness, etc. usually on the order of 10 minutes tho
if you use H2O2 + HCl in a glass dish instead, you can watch the process better and it's also a bit faster. also you never have to buy more, just aerate the solution and it's ready for round n+1

>>1732529
>LM317s
>being this cucked

>>1732530
photodiodes make decent light sensors. you have to change the current around a bit. PV solar cells are just great big photodiodes

>>1732600
SSTV is just an FM encoding standard for video
shorter version: it was basically a CRT in reverse. like core memory, the general idea was to push some energy into something in a controlled fashion and measure its response
https://en.wikipedia.org/wiki/Video_camera_tube#Vidicon
>muh fake moon landing
did you really have to ask something so easily web searchable in order to make a shitty flat-earth meme post? are you really going to keep committing the sins for which Epstein didn't kill himself for you?

I added a battery charge check to my simple circuit, will that work?
i don't want the buck converter to explode or something

the battery will be a 2 cell lipo (8V fully charged)

>>1732638
you need the X and Y commands to mirror your shit and draw it in a logical order
R1 and R2 could be an order bigger. the input resistance R(ain) = 100MΩ so you have a lot of room for less resistor divider current, so as not to drain the battery with your battery voltage sampling
actually, your battery pack might output up to 8.6V under some conditions. leave some margin when you choose R1 and R2 so that the PB3 input voltage is *always* between Vcc and ground, even during charging the battery, or while your buck is starting up, or just because your buck is made with 5% parts and kinda sucks as a regulator
don't use Vcc as a reference, if you set R1=3*R2 and use the 2.56V internal reference, you can read out the input voltage directly in 10mV increments

Could someone explain to me the point of R4 in this configuration please?
http://www.ti.com/lit/ds/symlink/lm5085.pdf
The explanation in the Reduced Ripple Configuration doesn't make much sense to me either because I don't see how R4 would effect the ripple voltage to the FB pin

>>1732692
I believe it is to make the control loop stable, like they do with LDO regulators. If the ESR of the capacitor is out of a certain range then Vout starts oscilating. I'm not sure tho, because they did not really explain R4 in the datasheet.

>>1732692
constant-on-time. R4 and Cout are meant to reduce output ripple. however, some minimum ripple is required for the converter to turn on. R4 increases the ESR of the output cap so that the converter can still see ripple. note that you'll need an R4 or several for the load if it has a significant amount of capacitance across the rails elsewhere
(if you want a clean output, this may not be the converter for you)

>>1732702
Well I would just use it to charge my phone. Could I just ignore R4?
You and the datasheet say R4 is necessary so that FB will get some ripple to be able to operate but I don't really see how R4 effects FB in any way.

>>1732704
that's quite a bit of converter just to charge a phone
no, you can't just ignore R4. if you replace it with a short, the output will be too smooth for the converter to oscillate
energy is passed from the input through the inductor in pulses. Cout accepts the energy and stores it to even out the output. FB is a function of the instantaneous output voltage. R4 reduces the capacitor's smoothing effect by resisting current into/out of Cout

>>1731739
Check out johnaudiotech YouTube channel, and freeze frame his ltspice model of one of his transistor amplifiers. Mimic his .trans and .op

LTspice... LT.... Linear.... no good for audio... LAMFAO.

>>1732767
>no good for audio
I wouldn't say it's bad, but I don't think it has the handy "headphone out" feature. It would be nice if you could do a live simulation with custom audio in (via some sort of media player) instead of a simple sine/square/triangle/saw, as well as audio out. If live is too much for spice, then a prerecorded one that outputs an audio file would be acceptable also.

>>1731876
You can make a joule thief to boost the voltage by some arbitrary amount, one transistor. Voltage upped by adding more “secondary” windings.

>>1732772
You can get audio input from files, and record the output. Maybe you could get the output and play it in a audio player. Maybe it's too much work but meh.

>>1732776
>https://www.electro-tech-online.com/threads/using-wav-files-as-input-and-output-in-ltspice.123395/
>https://electrostud.fandom.com/wiki/Using_WAVE_files_as_input/output_in_LTSpice
I too went through the phase of hating LTspice, but now that I have learned how to use it I enjoy it very much.

>>1731913
You probably want a transformer for isolation, yeah. If you look at a graphics card, it’s basically one chip (gpu) and it takes in 12v and has a dozen dc/dc voltage converters on it to down convert to 3.3 or 1.1 volts. At 100A.

>>1732288
Literally no legit reason to use CdS cells anymore... they even wear out over some years. Only for old equipment replacements and old designs (laziness)

Anyone know shit about the Apple II's composite video output? Or about using low-resolution composite video on a modern flat-panel TV?

I switched to 80-column mode and text was practically unreadable.

>>1732519
Soak until done. Not necessary to heat, but it makes it go faster.

You can plate your slobbering iron tip with it, too.

For the love of god though, don’t leave the lid off the ferric chloride bottle.

>>1732529
You can use lm317s as amplifiers.

>>1732794
colour demodulator no likey 80-column mode. traditionally you'd hook up a black & white TV.
with a modern flatscreen I guess you could feed into the Green of an RGB input.

>>1732802
Fuck. I could've sworn it looked fine in the 80s... but I was just a tot back then. Maybe I wasn't using 80-column that often.

Thanks.

I want to get started with electronics. I was looking at buying two of the books from the OP. I am wondering what is the best option to get the components I need. Was looking at something like pic related

>>1732808
ebay, ali express

>>1732802
There's not enough luma resolution in composite video for 80 column text. At the very least you need S-video. C64 supported it, but not Apple II.

>>1732833
Crazy idea, if your monitor has an S-video input, just hook up the video to the Luma pin and leave Chroma unused, the 80 column might work that way.

>>1732808
That looks like a decent starter set, get a 555 chip or two to go with it.
You might also want to cut the end off of a USB charger cable to power it from a cheap charger brick instead of burning batteries. Small and plentiful 5V power supplies was a cool thing I did not have as a kid in the late '70s. Get a strip of 0.1" pins, break off a couple, then solder the end of the wires to them so it will will stay on the breadboard. Note: do not connect an LED directly across 5V 1A, it may explode its top off.
FYI those dupont type wires may be convenient, but I think their pins break too easy. I'd rather just clip and strip 22ga solid wire.

>>1732833
>>1732834
Christ. I might just try to jerry-rig an RGB monitor output. I think I have an old mac monitor lying around someplace, and I think the 80-column cards have the headers for something like that on the board.

>>1732838
Apple II isn't RGB. It uses NTSC chroma phase shifting to do color. The IIgs has RGB support though. But even more important, good luck finding a monitor that can sync to 15KHz RGB, they're not common. If you could get that far, I guess you could just tie all three colors together.

And now I remember these fuckers. When you displayed color on them, you could see the color phase pattern bars, so they clearly had no low-pass chroma filter and worked fine with 80 column.

>>1732802
>>1732833
>>1732834
>>1732846
Am I gonna have the same problems with my Timex Sinclair 1000/ZX-81?
(once I get around to hacking together a wall wart that it'll take)

>>1732891
It doesn't do 80 columns, it doesn't even do 40. It was always meant to work with standard NTSC. As long as the image keeps to 256 or less pixels per scan line it's fine.

>>1732896
lel it doesn't even do color

>>1732901
Yeah, you should see the keyboard. It doesn't even really have tactile keys. I'm surprised my dad didn't throw it out years ago. I don't even have any tapes for it.

>>1732808
>only two of each sort of transistor
>not even complimentary pairs
Go for 50/100 packs of assorted transistors (NPN and PNP), capacitors (electrolytic and ceramic), LEDs, and diodes. Even more for resistors. Some zeners, linregs, trimpots, inductors, etc. might also be a good idea. For ICs, get some comparators and op-amps, they're the workhorses of analog circuits. Some MOSFETs aren't a bad idea if you plan on switching any sort of load.

>>1732834
You know, the last time I tried to take this thing out (like 7 years ago) I might've used a composite-to-S-video cable. I swear I used to have one of these.

No idea if 80-col worked or if I got color though. I think I'd just been trying to get some of dad's old files transferred and gave up when I realized how painfully slow the tape deck I/O was.

>>1732834
>>1732913
Son of a bitch. https://youtu.be/wQaouPzMSsI?t=506 (skip to about 8:30; he tied the chroma and luma together rather than leaving the chroma unused). When he goes into 80-col mode I just about came.

The weird vertical banding in graphics modes is actually kind of aesthetic too.

>>1732808
on a similar note, how good the recommended electronics kit bundle for the Platt book ?

>>1732906
I had one as a kid. quite an object lesson in how minimal a computer could really be

>>1732911
>buy a bunch of random-valued crap he may not need or ever use without even knowing what the values mean
dude, stop

>>1732927
no idea. probably pretty okay if you follow the Platt book. as usual you can probably get the components from LCSC for less than the faker bundle

>>1732953
>random-valued
Ok at least a grab-bag and/or powers of 10 of resistors and a bunch of powers of 10 of ceramics and electrolytics. Best thing I bought recently was 100 of 10kΩ, 100kΩ, and 1MΩ resistors each. Probably shoulda got 1kΩ and 100Ω too, which I didn't think I'd use much but for setting minimum gains and covering more orders with passive filters they're actually pretty good.

I wish you could buy assorted packs that are only 100, 1k, 10k, 100k, 1M or 10p, 100p, 1n, 10n, 100n.

>>1732997
>I wish you could buy assorted packs that are only 100, 1k, 10k, 100k, 1M or 10p, 100p, 1n, 10n, 100n.
Buy large volumes, repackage, and then sell on eBay.

>>1732999
I don't live in the states or anywhere with more than a few million citizens, so it's just not worth it.

>>1733002
Should still look at what shipping costs would be. You might be able to send resistors the same way you would letters. Might be stupidly fucking cheap.

>>1733009
>the same way you would letters
Oh yeah. Custom PCBs could also be sent like that, or aluminium solder paste masks, perhaps even specialty flat populated board circuits.

>>1732519
If you want to make quick pcb prototypes i recommend you get cnc3018 from chinks
fecl3ing pcbs is nasty shit and doesn't belong in modern hobbyist's lab
not to mention you are fucked when you want to to have precise tht holes and via holes in your shit

>>1733048
yeah, modern hobbyists use CuCl2
>tht holes
doesn't belong in modern hobbyist's lab

>>1733054
>doesn't belong in modern hobbyist's lab
Especially since plated tht holes are a nightmare to diy. But what sort of board has absolutely no holes at all? Arguably you could avoid using both sides of a board with 0Ω links. I know IC sockets can be turned into SMT versions just by bending the flat feet, and bent pin headers can be used as SMT parts, but what about when you want a transistor or linreg to dissipate a watt or two? Not like you'll have dozens of vias to conduct heat away from the parts. Bend the legs of a TO-220 to mount it on a heat sink? Sounds like asking for a delam. Glue it upside-down? And what about other connectors that need some sort of structural support, like BNC, micro USB, banana, etc?
Having a router also provides the convenience of cutting out the board outline, and making isolation slots and cutouts for military-style dual inline flatpacks and the like. As far as SMD prototyping goes, I think a router beats an etch setup for these reasons. Though personally I'd want both CNC routing for the holes and cutouts and such and a chemical setup for my solder mask and etch resist along with electroless THT plating, with an LCD programmable lightbox that can also expose silkscreens.

>>1733054
>doesn't belong in modern hobbyist's lab
Don't make me laugh.

How exactly do you plan Where exactly do you plan on getting something like an SMD linear voltage regulator that needs to dissipate like 5W and thus needs a huge heatsink? sorry kiddo, tht only
same with DIP sockets for chips, maybe you want to use say attiny in your project and be able to easily remove the attiny and reprogram it so you use a THT DIP8 socket
many cases where THT has a place in the modern world of electronics

(I'm the guy who invented pcbs so i know what i'm talking about)

for me lads, it's dead bug mounting

>>1733079
free floating experiment

>>1732997
I'm trying to imagine what I would do with even one 1M resistor, much less 100 of them. It depends on what you're building, if you're doing analog circuits, I guess, but with digital circuits, I can't remember using above 10K since forever.
For me, small values are useful, mostly 1K-5K for pull-ups, and 220-680 for LEDs depending on supply voltage.

>>1732911
>comparators and op-amps
they are the same thing

>>1732808
I started with a $15 "elegoo" kit. My first project was a transister oscillater for led's, which I then used as the RCLK/SRCLK to power blinking light sequences using a shift register (not a simple task, for a beginner).
It'll teach you everything you need to know about breadboards, digital I/O, and parasitics, at the least.

>>1733113
Comparators are faster at being comparators

>>1733113
>they are the same thing
no doubt

>>1733059
I just scrape copper of copperclad with a xknife. Zero holes. Git good zoomer

>>1733113
They have much in common, but there's a reason why comparators exist as a distinct product. Specifically, op-amps don't like being driven into saturation, whereas that's expected behaviour for a comparator.

>>1733076
>I'm the guy who invented pcbs
What exactly causes vias to form a pink ring around them?

>>1733236
It happens when they reach puberty and become fertile.

>>1733236
>>1733242
then the blue-green color vias must be the male ones

>>1733236
not him but from what I remember it's actually diffused copper in the FR4
might be caused by unsharp drill bits or a bad etching process but idk

>Use a DC PSU rated at 1.9A for a 17A system
>Room now smells of burnt plastic

If you connect DC + wire from phone charger adapter to earth ground in the outlet it won't do anything because it's isolated right?

>>1733076
>removing an ATtiny to reprogram it
UPDI and choose your pins carefully, summer child

>>1733108
>10k pullups
sad

>>1733059
>sockets
lel
>power dissipation
a fair point for some things, but I'd consider pigtails
>mechanical
if you need mechanical unity between a heat sink etc. and a board, drill holes or use a pigtail
why are you using BNC instead of edge-launch SMA anyway :^)
>As far as SMD prototyping goes, I think a router beats an etch setup for these reasons
I disagree about "for those reasons" but you might be right about a router more reliably and cleanly handling finer pitches. 0.5mm pitch does stretch the limits of a home etch system, never mind gouge-and-peel
otoh I don't see a whole lot of wisdom in that level of home lab except for the most active hobbyist. might be worth joining a makerspace at that rate

>>1733260
makes perfect sense that a wet chemical process would reach into a mechanically roughed-up fiberglass medium

>>1733274
should be fine but be careful if you connect that gear up to something else that expects V- at earth potential

>>1733274
Yep. There should be zero voltage between neutral and ground. If you see sparks, I'd recommend replacing the charger.

>>1733286
>>10k pullups
>can't read
sad

>>1733288
> 1-5k pullups
2-10x as sad

>>1733108
Low q-current voltage dividers and feedback circuits = lots of 1M resistors. I tend to use 100ks as pullups unless the datasheet tells me not to (i.e. all fucking comparators under the sun).

>>1733113
If using negative feedback, an op-amp will have better characteristics. If not using any feedback, or using positive feedback, a comparator is more optimised to the task.

>>1733127
I didn't even know that existed, interesting part.

>>1733286
>>sockets
>lel
For audio stuff you often want to swap op-amps for other op-amps
>0.5mm pitch does stretch the limits of a home etch system
But those 0.3mm pitch boards are done by industrial etch systems, right? I wonder if they add leaveners or other additives to their chemical soup, or perhaps thinner etch resists. For a homegamer, I bet you could pre-emptively make the traces themselves thinner or thicker in the gerber file (or afterwards in the svg) such that the process of exposing and etching results in the correct width of copper.

>>1733286
>makes perfect sense that a wet chemical process would reach into a mechanically roughed-up fiberglass medium
Then explain why not every via has a pink ring

>>1733338
>But those 0.3mm pitch boards are done by industrial etch systems, right
the operative word is systems, that monitor and correct themselves, and as a general rule don't have to deal with fingerprints or other variables
like all wet chemistry, it's a matter of knowing what you're working with. if you're reusing a bath of partially spent FeCl3 from the last run without any analysis or treatment, the terrorists have already won. (one of the advantages of CuCl2 baths is that they can be titrated and adjusted very easily)

>>1733339
lucky drill hit, sharp bit, drill didn't hit the glass cloth in just the wrong place, drill polished the hole on the way out
just accept it as god's will and follow Her design rules

>>1733384
I use persulfate, what's the deal with Cu2Cl? Is it one of the refreshable etchants? Couldn't you refresh it electrochemically if you used CuSO4 instead?

>>1733339
those are the roasties

>>1733389
why, yes, it is. CuCl2 can be regenerated by bubbling air thru it
http://jimlaurwilliams.org/projects/seychellePaper/index.html

for the love of fuck, where can I get one of these

>>1733433
>bubbling air
Oh, that's simpler than having to use H2O2. But I see it still uses up H+ ions in the regeneration process. That's why I wonder about using CuSO4. In this case, you'd get:
CuSO4(aq) + Cu(s) > Cu2SO4(aq)
when etching, and:
Cu2SO4(aq) > CuSO4(aq) + Cu(s)
when regenerating with a copper and graphite electrode.

>>1733448
Why do you want one? I bet the digital storage method is old enough that it's difficult to export the data from the scope. I hope it has at least 2 channels.

>>1733451
just look how small it is, it's adorable

is there a specific name for the actual incandescent bulb in a traditional christmas light?

Do Lightning and USB-C handle analog audio signal differently? I'd like to make a
[3.5mm male + USB-A male] -> [USB-A female]
so I can plug in a USB-A to Lightning or USB-C cable to charge and AUX a phone. Is it (easily) possible, or do I have to get a dedicated Lightning/USB-C dongle?

>>1733482
pic related

>>1733482
>analog audio signal
Do they even carry analog audio at all? I always assumed that they required an external DAC shoved into those dongles.

>>1733451
ahhh. I see your point, that the H2SO4 version could be less exacting to regen. interesting. if you have H2SO4 readily available to you it might be worth a science. that's about the limit of what I would like to keep around the house, personally

>>1733453
holy kawaii batman

>>1733482
>Lightning
what makes you think Apple's gonna recognize your kit, let alone talk to it?

>>1733512
>H2SO4
No, my point was twofold:
Firstly, that you wouldn't need to add any extra acid if you refreshed the etchant electrolytically rather than with atmospheric oxygen (avoiding the O2 + 4H+ = 2H2O that uses up the H+ ions)
Secondly, using SO4(2-) ions instead of Cl- ions would allow you to use electrolysis without fear of oxidising chlorine gas, an event which would require pH correction to make the solution more acidic.
Ideally you wouldn't need any acid at all, you could start with just some copper sulfate solution.

But I see two problems with this concept:
Firstly, it will likely require some pH engineering in order to get it to reverse at all and this pH may not be suitable for fast etching.
Secondly, oxygen will still dissolve into the solution so long as the solution is present in the air, oxidising any Cu(1+) ions and increasing the solution's pH. With any luck, the process of electrolysis will liberate some oxygen gas and drop the pH again without any additives, but that would be rather optimistic.

>>1733542

I just use (NH4)2S2O8... Ammonium persulfate. You just mix it with water and then it turn pretty blue color when etching.

>>1731520

>>1733546
That's what I use too. But according to that link in >>1733433 it's slow as shit, and needs to be dumped sooner or later. No clue what they were thinking, using ammonium, a heavy and somewhat weakly dissociating ion, as their cation.

>>1733549

Still easier, cheaper and faster than having china make 'em for you. Like how fast do you need it to be? Heat it up if you want it to go faster, or just jack off while waiting so you're distracted for 10 or 20 minutes.

On the plus side, you won't be so frustrated and you will hardly ever over-etch the board. It's clear and you can see it as well.

>>1733542
ok. but H+ ions are cheap. almost certainly cheaper than electrolysis
neat idea tho

>>1733553
trouble is, the more boards you make in one sitting, the longer the timer runs

>>1733569
>almost certainly cheaper than electrolysis
What, you mean cheaper than the electricity? Because the copper is basically renewable, in fact I get more copper out of the etchant from what comes out of the etched boards.

Does anyone have experience with the ADNS-2700 optical mouse sensor? It should function as a USB mouse on its own.

I have a bunch but I can't get them to work. Windows fails to enumerate the device after connecting the VDD, GND, D- and D+ pins to a USB cable (Get Device Descriptor fail).

Any ideas how to debug this? I don't have an oscilloscope unfortunately.

How am i measuring 230V AC between ground and live wire when the live wire is physically disconnected by a switch? and the switch is not broken

>>1733610

phantom induced voltage. put a load across that 230V, even a slight load, like a 1M resistor, and see if disappear. or use an analog meter which will show the true voltage.

>>1733614
I am getting tired of this casper shit every time i try to do something with electronics.
Why do i even bother measuring shit with my multimeter when it always shows shit is at 230V and then it becomes a russian roulette. I touch it and either nothing happens or my pacemaker blows up

>>1733626
It is caused by poor grounding. If you draw a circuit of what you think your measuring maybe it would help.

>>1733448
If you ever get one please do a photoshoot of kittens and other cute little animals snuggling with it.

>>1733581
HCl is $6/gal where I live
>I get over-unity copper out of the solution
ok. and then what do you do with it? best application I can think of is to grind it into dust and keep it in a catsup bottle in case I have to put out a lithium fire

>>1733602
show schematic

>>1733614
>muh true analog voltage
no it's not, it's just a load resistor. stop memeing

>>1733274
Switching PSUs typically aren't isolated, but have capacitors (either class Y or 2x class X in series) between DC ground and both live and neutral. This is to prevent high-frequency switching noise being conducted back into the mains supply (which will act as an antenna and radiate it).

If you put a scope probe on the DC ground, you'll see half mains voltage; you can often feel a slight tingle if you touch the DC ground connection. The impedance of the capacitors is high enough that you can treat the supply as isolated for most practical purposes, although it can be an issue for signal connections to very high impedance inputs.

>>1733751
>show schematic

>>1733834
power supply bypass caps aren't optional

>>1731520
Is a "2N2222" transistor the same as an "MPS2222A" transistor? I went to a shop, showed the guy the identifier, and he handed me the other one.

>>1733866
RTFD

>>1733867
you useless cunt

>>1733866
>can't into a datasheet
>calls others useless
pure pottery

imagine being unable to rtfd

>>1733866
It's not even a real 2N2222 transistor unless it's in a metal can. If it's in a plastic case it's a PN2222.

The grounding trick worked and the traces milled perfectly.
I probably have all kinds of lung cancer from breathing in all this powdered fiber glass

1mm and .5mm traces were absolutely no problem and look nice and crisp
A .25mm trace got fucked up tho and is interrupted, but i think this can be solved by setting a slightly larder diameter for my vbit, since i set it to exactly .2mm so it had no tolerances really, with some adjusting i think i will be able to do .25mm

>>1733907
Fuck. I just realized the dip8 is soldered from the other side so the footprint has to be flipped around
Oh well at least i have an excuse to do more milling

>>1733626
Consider whipping up multimeter leads with ~270kΩ parallel resistance for these cases. It's about the smallest resistor that can withstand mains across it at under 1/4W.

>>1733751
>what do you do with it
Electroplate it back onto copperclad for high-power traces, perhaps electroplating onto aluminium or nichrome or kanthal or something so I can wet solder to it, assuming that it won't get above 200C in normal usage.

>>1733907
>lung cancer
Nah, just silicosis.

>>1733909
>dip8 is soldered from the other side so the footprint has to be flipped around
Yeah, it's pretty easy to make that mistake, done that plenty often myself.

Want to fix a ~10 year old Samsung LCD tv which mostly works but the picture is solarized. There was a time when hitting the chassis forcefully would somehow nudge it into displaying the image properly.
I have no experience but I have a multimeter, a soldering iron and flux should it be needed.
I was thinking of checking out
>Geier, How to Diagnose & Fix Everything Electronic
from the OP.

Any tips?

>>1733934
Caps. It's almost always the caps.

Rec me some 5V medium current MOSFETs to jam in low voltage PWM circuits
preferably complimentary ones

>>1733907
lel, the wages of THT cuckery

>>1733983
freedom isn't free and neither are MOSFETs
>medium current
any reason I shouldn't smack the shit out of you for not being numeric enough?

>>1733934
>solarized
posterized? check the flat cables and connectors first. THEN caps

>>1733922
> ~270kΩ
>withstand mains across it at under 1/4W.

Huh... that's an interesting factoid to know. Thanks, anon.
(I assume you mean the 120v yankee mains, not the limey mains)

>>1734075
By medium

>>1734078
Nah that's 240V, because he was talking about 240V. For 120V it's 68kΩ. V^2/0.25 = V^2*4 = R

>>1734078
R=V^2/W. 240V->230.4kΩ, 120V->57.6kΩ

>>1733893
> metal can real 2N2222

I believe ON semiconductor inherited Motorola's transistor fabs, so those are probably as "real" as you're gonna get (most of the 2N2222s I have are from the 70s and Motorola)

>>1734082
I'd use two 130kΩ in parallel to minimize the dissipation. If you run the resistor too close to its max power rating it's gonna get hot, once you derate it for temperature you'll find it's exceeding its power rating and it'll eventually fail prematurely. Two 130k in parallel gives you an effective 65k resistance and each device will only dissipate around 1/8W so they should stay cooler.

>>1733569
>H+ ions are cheap

I just put my PCBs in a tray and vomit on it. The HCl in stomach acid is virtually free!

>>1734089
The point is that the full spec power dissipation of 2N2222 needs the metal can, the plastic version can't do it.
inb4
>using a 2N2222 as a power transistor

>>1734099
>power dissipation
>needs the metal can

problem is, metal-can 2n2222s are crazy expensive; also assume that like the 2n3055's all the cheap e-bay ones are fake.

Do the plastic and metal can 2n2222s have the same die/chip inside?
maybe you could grind away most of the low-thermal conductivity plastic and then attach it to something (somehow) that would sink heat away faster.

>>1734103
Or, how about if you need a transistor that needs to dissipate heat you just get something in a TO-220 package instead. Not the 2N2222 as it doesn't come in that package type but it's a jellybean transistor, you can find some suitable replacement easily.

>>1734075
Shit, by medium I meant something like 25A or whatever. But I'll likely never use more than 10A. What matters more is a relatively low threshold voltage (i.e. never gets above 3V) and a gate charge that's low enough that I can run it at a few 10s of kHz without issue.

>>1734106
>by medium I meant something like 25A or whatever.

25A will make an alligator lead glow and break in 5 seconds. it is therefore not a medium current, but a high current.

>>1734113
>high current
I thought high currents were the 50A+ ones I see on digikey or wherever
20A seems to be almost a minimum for common power mosfets

>>1734106
What is your power supply in your circuit?

>>1734125
I said 5V, but that's more of a minimum, I'm likely to use it up to 12V, and possibly up to 36V. Having a V_g-th low enough to use 3.3V logic would be kinda nice, but I doubt I'll ever use a 3.3V rail without a higher voltage rail for switching high-power loads, in which case I can simply use an NPN or two to switch that higher-voltage rail to the FET.

>>1734113
orange vests are for pussies

>>1734106
physics aren't especially complementary when it comes to MOSFETs. Rds(on) * gate charge on the p-channel will be ~3x more than the n-channel, all else equal
tons upon tons of options out there with <0.1Ω Rds(on) specified at 5V. the Si7456's specs are impressive

>>1734096
this anon's proton pump fucks

>>1734200
>Si7456
>obscure SMT package
>new product
>having to scrounge up a free 50 burgerbux in order to not pay out the ass in shipping from digikey or wherever

Ok, new plan. My priorities are getting shit cheap from ebay or ali or wherever, and having it be a thru-hole , probably TO-220. When I need to make an SMT board I can always just trim and bend the tab and leads of a TO-220 into a D2PAK. Don't know why people bother buying specialty SMT power transistors anyhow.

Accidentally posted this in /mcg/ lol
>MV209's are no longer officially sold
>Ebay only has them for $2 apiece
What is the go-to solution for varicaps these days? Just pick any random part that's cheap and works?

I am thinking of designing a a small buck voncerter and I'm not sure if a small sot23 mosfet would be enough
I'm pretty sure to calculate the power dissipation there are more than just the "Q*U*Fs" since it only calculates the power needed to switch the mosfet on and off.
What am I missing? I know about Rds on multiplied with the current but that probably only applies to a static state of the mosfet

what's safe to buy from chinkland? I'm thinking of picking up 50x 555s for stupid cheap. Am I right in assuming the worst that could happen is it pulling more watts?

>>1734293
I mean, they could also be DOA. But for something so fucking common as a 555 I doubt they're going to be anything abnormal if they work.

>>1734293
Uh.. Quite safe. And when I got problems like mail fuckups I always get refunded.

>>1733907
second attempt, i tuned the settings a bit and this time even the 0.25mm traces made it intact, i also used a .8mm drill to have more copper left around the holes to solder, i continuity checked everything and there are no shorts or interrupts, yay
Now i will try to solder it which will be a nightmare because soldering smds without a solder mask will be probably near impossible, without shorting them to the ground plane. Well the small ones, the big ones will be ez

>>1734306
Maybe next time mill the area under the SMD parts? For now you could try a knife to cut and scrape out the extra copper.

>>1734146
No no no, what IS your power supply? Not the supply voltage but the supply topology that will deliver your 5V 25A? Switching converter, battery, unregulated linear supply, regulated linear supply?

>>1733448
ebay

I am retarded what are these?
Can I cut the black part to get individual pins?

>>1734306
the back side of a box opener can be used to carve up the fiberglass to a greater depth and stop solder from bridging. And if a bridge does form it's easier to remove it with a soldering iron than witha knife or mill.

>>1734391
"yes"

>>1734391
I call them .1" pin headers.
Yeah, you can split them.

>>1734391
they are pins, you can snap them apart but you might want to use a box cutter knife to ensure a good break

>>1734241
>go-to solution for varicaps
SDR

Party time.

You are hosting a party and inviting all the neighborhood electrons to attend. So you broadcast "PARTY AT MA HOUZE YOLO #SWAG!". I.e. you apply a positive voltage at the gate with respect to the source. Given that your neighbors are some distance away (next door), your broadcast isn't loud enough (You are below Vgs(th)) Once you yell your invite loud enough (i.e. Have Vgs at Vgs(th)) your next door neighbors hear and come party with you.

Vgs(th) is the voltage at which the mosfet channel begins to conduct. At this voltage, a positive voltage, it creates an electric field, which attract electrons (since our applied voltage is positive, so positive charges on gate). These accumulated electrons near the gate, form a bridge between the source and the drain (which are both n type). Now you have a "continuous" n type path from source to drain.

You only attracted your next door neighbors, so your party is kinda lame. How do you get more people? You broadcast louder (Increase the range of your electric field - increase your Vgs).

How hot is the commutator supposed to get?

Also, aren’t brushed motors running off DC? So where is it getting switched from AC to DC?

>>1734416

Hot enough to burn you, and hot enough to melt ordinary plastics.

The DC is converted to "AC" -- well pulsating DC... close enough -- by the commutator (those gaps in the copper commutator connect the brushes to separate coils)

>>1734421
Ah ok. The thing started crapping out on me, nothing seems loose inside. Maybe it’s that outdoor outlet too, feels like that thing is a little bit loose and I don’t use it much.

So, what would happen if I hooked this thing up to 18VDC?

>>1734416
It's a universal motor, it seems

>>1734423
>>1734426
nothin

>>1734423
> 18vdc

Yeah, that's doubtful. You're gonna need something higher. The "electronics 'n more" guy just did a video about powering stuff with DC on universal motors, but he was using a much higher voltage.

>>1734412
Soulless

>>1734241
good old 5mm red LEDs http://www.hanssummers.com/varicap/varicapled.html

>>1734265
Rds(on) applies to both peak and steady states. most switching MOSFETs will specify maxima for both

>>1734293
>chinkland
>safe
le mayo
you could find re-marked LM358s in that batch if you're unlucky
you'll almost certainly be receiving pulls

>>1734306
>soldering smds without a solder mask will be probably near impossible
position them carefully and you'll be just fine. but next time carve a bit wider channel around their pads
nice neck down

>>1734415
anon shows us what the n in n-channel stands for

>>1734204
>SO-8 sized DFN is obscure
ok boomer
>/ohm/ is /csg/
if you're going to limit yoursel to chinkshit, your best strategy is to find a chink, flip through all 80 pages of their store, and look up part numbers that appear to be in the right case. you will then come out the other side with a better knowledge of what is being recycled this year
>don't know why
because they're not shit designers and they're not boomers. if you're dissipating more than a watt or two in your switch, your design has a problem, and that problem is you
millions of these things go into notebooks, tablets, and all sorts of other devices for load switching, power supply driving, and other purposes.
anyway since you've sat thru or skimmed over this much of my abuse, look for 30N06 part numbers (30 A continuous at 25°C, 06 * 10 = 60V d-s breakdown) and you may find a series that suits you in whatever package you like, with sufficiently low Rds(on) that you won't even need a damn heatsink if you drive it properly. Fairchild/ON's FQD30N06, maybe

>>1734437
>le mayo
Not sure I get this gag. Is it a play on 沒有?

>>1734480
lmao

http://www.ti.com/lit/ds/symlink/lm5085.pdf
So I'm still messing around with this DCDC buck converter and I'm trying to choose a mosfet.

In the datasheet on page 15 it says
>"The gate-to-sourcevoltage (VGS) normally provided to the PFET is 7.7V for VIN greater than 8.5V "
Doesn't the PFET need a negative VGS?
If VGS = 7,7V at Vin = 8,5V wouldn't that mean that VS = 8,5V and VG = 16,2V because VG - VS = VGS? This wouldnt make any sense.

For the example circuit they are using a Si7465DP mosfet from vishay.
That one has a VGS of +/-20V max.
Since their maximum Vin is 55V VS would be 55V too.
Doesn't that mean the Gate Voltage would need to be high too? Otherwise the PFET would break.
From what I can tell from the functional block diagram the Gate of the PFET is either VIN or VCC (so in this case 55V or 7,7V, therefore it would break)

>>1734507
oh
jeez

>>1734241
1N4000 series diodes

>>1734521
Ok I think I answered my own first question.
But I still dont know what they were thinking with a low VGS of +/-20V if the input voltage is 55V.

>In Transit, Arriving Late
mother ass fuckers, there goes my weekend

>>1734521
they weren't thinking about fussing about signs, that's all. not many people do fuss about the signs outside of MOSFET datasheets

>>1734528
The 1n4001's I'm using are shit though. Nothing happens from 0-4v, then at 5V the Q of the LC tank just drops. The resonant frequency increases maaybe just a little.

>>1734609
I guess try LEDs like the other guy said
or even junctions from transistors

>>1734241
Get one of pic related and attach a teeny tiny servo motor controlled by an arduino onto it.

>>1734609
from that dude's website
>1n4001
the higher-voltage, higher-current diodes seem to be a bit closer to the capacitance range you want
but if you need great Q, probably better just pick a varicap

>>1734813
>cucks actually do this

Is there a difference between these inverters?

>>1734854

top one is straight, bottom one takes it from behind. they both invert equally otherwise.

>>1734867
>>1734854
ie. the top one is a little less gay

>>1734306
Another day, another attempt.
I am getting pretty comfy with the workflow. This time i set even bigger tolerances to account for tool runout and set up clearing paths that remove extra copper around the pads so it will be easier to solder smds without having solder bridges connecting onto the ground plane
Currently i'm waiting for the autolevelling process to finish, can't wait to see how the attempt #3 turns out

>>1734902
Final version
It turned out great. Traces are nice and uniform, with plenty of space for soldering wihout any dangers of shorts
And now to hope i will be able to solder stuff properly with my shitty flat tip

>>1734911
Fuck this will be so hard.
I purchased the largest sized smds i could find and i was still shocked at how tiny they are.
Why can't they sell tht sized smds god damn it, soldering this microshit is so frustrating

>>1735014
I used to work at a place where they had pick and place machines. One was literally 10m away from my desk. The guy who vacuumed them out was based and would put a bag of the stuff on my desk every now and then. Then I'd take it home and sort through it. Some of that shit was almost sand grain sized, I called it "grit".
Yeah, this is why a production-quality PnP machine needs visual sensors to verify that the part actually got placed. I was surprised how many parts fell down into the bowels of the machine.

>>1735017
I miss the times when everything was made from huge THTs that even a retard could work with easily. An average skilled hobbyist could output products of same quality as a factory could
But today? Any industry standard components are basically atom sized and thts or larger smds are just some hobbyist niché
I'm sure that pretty soon we won't be able to make stuff at home that is even like 5% as good as factory stuff. Like even today, you can't make anything that is even CLOSE to shit like an intel CPU
But then again, maybe some time in the future we will have really good 3d printers that will just print a finished pcb with components already on it, from some sillicon powder or some shit

>>1732174
>do you just recognize the forms of common component groups and their functions
Pretty much. You start this way.
If there's chips, the absolute first thing I do is look up their datasheets. You can tell a lot about a circuit just by looking at the chips and what components are around them, much of the time without even needing to know exactly how they're connected together.

>>1734391
You can also cut the female version of this but you will always lose 1 pin. Just been my experience.

>>1734241
What capacitance do you need?
Here is bb910. Plenty of BBxxx on aliexpress.
bb112 is ~400pF
You just need to search for varactors not varicaps :)
https://www.aliexpress.com/item/32904746572.html
>20pcs 76c

>>1735044
posted wrong one

>>1735026
We still will be able to do "things" but the gap is widening for sure. Still, we do have access to amazing technologies compared to 10-20-30 years ago that are also cheap and taken for granted. But the bleeding edge stuff is getting exponentially more complicated and integrated etc. But this has been going on forever as ICs are getting more and more high level and integrate more and more functionality etc.
> anything that is even CLOSE to shit like an intel CPU
So, case in point: we do have access to super cheap microcontrollers. and the best of them are getting pretty fast. Of course they are not comparable to intel but what did we have 30 years ago at the hobby level? z80 homebrew?

>>1735051
>So, case in point: we do have access to super cheap microcontrollers.
Not only that, but it's usually easy to find them already mounted on semi-standard boards, whether it's an ST Nucleo or a /csg/ Blue Pill. Now if only someone could reach out back in time to strangle the guy who put that 0.05" offset into the Arduino shield layout...

>>1732174
I'm not accomplished, just a 8th semester EE with electronics as a hobby. But after a while you really start don't caring about the values of resistors and bias points and stuff. You need to understand what the fuck is going on, and then go to the values and biases for a indepth analysis. I try to think of little blocks with certain functions. Like hierarchical sheets in a project. Sometimes you need to actually do math and circuit laws tho. (Specially in goddamn power electronics)

>>1735026
You are vastly overestimating the amount of skill required to work with small SMT parts.
An average skilled hobbyist today can absolutely produce PCBs of the same quality as a factory can. It just takes you a little longer.
On digikey, mouser, etc you have access to the same components as industrial professionals. Free software like kicad can do most of the design stuff that big software packages can, and PCB fabricators like JLC are user-friendly and sell small quantities for low prices.
There are reflow soldering ovens available for cheap, like the chinese T962 or the open-source controleo3 which are ~$200 or less, and you don't even need one. You can do half decent reflow with a hot plate.
Hand placement isn't as hard as it looks, with a magnifier and sharp tweezers. stereo microscope is god-tier but not required.
This is the age of SMT, embrace it. it's not as scary as it looks. and there's a lot of nice chips you can only get in SMT packages.

Somebody drop some thermocouple advice on me.
I have some diode modules that have a thermocouple embedded in the heat sink. The cold junction is on a PCB attached to the diode and the thermocouple connections are broken out to a pin header.
Naturally I tried setting my multimeter to temperature mode and hooking it up to that pin header, but it would only read around 20 degrees and didn't change much as the heat sink temperature changed.
What do I need to do to make this configuration work? I'm assuming the multimeter doesn't have any way to do cold junction compensation, and it just assumes the cold junction will be at room temperature. That assumption is just fine here, so why wouldn't it be reading correctly?
I've only used thermocouples hooked directly to the meter or to a module, I don't know how having pin headers and test leads in the circuit might be changing things.

>>1734876
an acceptor site's an acceptor side, said Vanilla ICE

>>1735014
are those 1206 or 0805? 1206 are on the same order of size as a 0.1" DIP, shouldn't be hard. you just need to add a toothpick to your arsenal to hold them down

>>1735026
ok boomer retard

>>1735056
>Now if only someone could reach out back in time to strangle the guy who put the Arduino
if you needed yet another reason to dismiss duinoshit as shit, there you go

>>1735113
this anon gets it
it takes a little bit of skill (ie knowing some tricks) and a little bit of tooling. the 1.6mm screwdriver tip isn't an all-access pass anymore. sorry not sorry

>>1735117
>the 1.6mm screwdriver tip isn't an all-access pass anymore
ha, I actually do almost all my soldering with a 1.5mm tip. Even tiny stuff. I don't go smaller unless I need the reach. The size of the tip isn't as important as how well you can control it.

I made this little LED star today. It's built on a piece of wood board and has an SO16-DIP16 adapter "PCB". It's a constant-current source made out of one resistor and a single ULN2003A. Currently this thing uses just 5 of its 6 constant current channels.

>>1735142
Detail view

>>1735142
>>1735143
That's fun!

>>1735142
>>1735143
Nice

>>1735120
for me, the 1mm bevel's got the right combination of mass and maneuverability for general soldering. suits everything from 0402s to 2.54mm pin headers. leaves a bit to be desired against ground planes tho. at some point you just need 2 irons. thinking seriously about le ebin TS-100 maymay or its successors

>>1735143
5/5 breddy gud

anons, how many SMD solder joints do you usually see before you wish you'd gotten a stencil made?

>>1735014
>Why can't they sell tht sized smds
They do, they're called MELFs.

>>1735193
>at some point you just need 2 irons
you do that? When I get to that point with ground planes and PTH I use the hot air gun, sometimes on the opposite side of the board.
I often think a tweezer iron would be nice, but every time I think I need one, I get the job done with hot air instead.

The TS100 is pretty great actually, I keep it with a battery pack and it goes anywhere. It definitely replaced my butane iron for all applications and I do almost all my general soldering with it now too. I can't seem to find a chisel/screwdriver tip smaller than 3mm though, that's my big complaint. But I just recently did some 0402 reworks with that tip and did pretty good.

>>1735297
frankly I'm mostly looking for a smaller iron. 936 clones are pleasantly cheap but I like to see what I'm doing while I work. with a 22cm tip-to-strain-relief distance I have to choose one or the other. sadly a Pace ADS200 isn't in the christmas budget

>>1735297
>I keep it with a battery pack and it goes anywhere
I too like to solder smds on long bus rides

Retard here, what is the difference between switching and linear power supply?

>>1735396
both have some sort of valve to pass current through to the load. switchers turn the valve on or off hard. using a storage tank of one or another sort to maintain even output current flow against those surges. linears turn the valve somewhere in between and create a variable resistance between in and out, dissipating that voltage * current = power as heat. switchers store the power in magnetic flux which can be readily converted to/from current and therefore lose less power. switchers also spend less time in that middle zone and don't produce as much "friction" overall, therefore run cooler
any particular questions?

>>1735396
Switching power supplies means buck or boost (or buck-boost). Flyback converters are sort of covered by those but sort of not. They all work by storing energy in an inductor (E = 0.5*L*I^2) and extracting it (via a diode) into a capacitor. This is theoretically lossless, but in reality there are switching losses and losses due to ESR.

A linear power supply is basically a voltage controlled resistor with negative feedback, set such that if the voltage at the output goes down by even a tiny amount the resistance will decrease a bunch, and vice-versa. The result being, the voltage at the output will always be approximately equal. Problem is, variable resistors dissipate a power P = I^2*R, or P = ∆V^2/R (where ∆V = Vsource - Voutput), so you have a much lower maximum theoretical efficiency. These are however much less noisy, both for in-circuit ripple and transmitted/radiated noise, so they're more ideal for extreme hi-fi, scientific apparatus, etc.

It's somewhat common, within mid/high-end power supplies where noise matters (from benchtop variables to specialised internal models) to have a switching supply most of the way and a linear one for the last few volts or so. This way you get the efficiency (and size) benefits of a switching supply, but the linearity and ripple rejection of a linear supply also.

For a mains > low-voltage DC power supply, there's effectively no room for a linear supply for all but the lowest of current applications (e.g. mains to low-voltage switching controllers). So switching power supplies are the most common here. But noteworthy are the transformer-based power supplies of yesteryear, which are still the workhorses in high-current applications like microwave ovens, (some?) welders, and some audio supplies where having a multitap transformer was easier than two seperate line switching converters to get a split-rail. Also noteworthy are capacitive droppers, which are decent enough to be used in modern LED light bulb PSUs.

>>1735401
>>1735404
Thanks, i see, that is exactly what i wanted to know

>>1731520
Guys my tv started to act weird.
Everything on it works but sometimes I need to power cycle (unplug/plug from mains) it multiple times in order to show picture.
If I dont power cycle multiple times the tv boots,shows picture and clock and then the screen goes blank and the status led says the tv is on and operating fine.(though there is no audio or picture).
Once I go through the "required" steps the thing works flawlessly as it always had.
Its a dumb tv from few years back so Google turned out nothing.
If anyone has ideas I want to hear them.

>>1735412
Looks like some caps died
Open it up and check if they have hemorrhoids on top

>>1735413
I was suspecting this but "its never the caps" meme got to me.
Gonna open it later today and post pictures.
Thanks anon.

>>1735414
>"its never the caps" meme
jesus why would such a meme exist?

Can anyone decipher this circuit diagram? In particular the parts marked with red. It's supposed to be a fuel level emulator for cars with LPG.

>>1735461
the thing that looks like a coil pretty clearly says "12V, 1030 Ohm".

The thing on the left looks almost like it says "51k".

Maybe find a better diagram. Idfk what's even going on there.

>>1735463
Sure it says 12V, 1030 Ohm, but what inductors are rated as 1030 Ohm? Inductors are normally rated in henry. The resistor, too, why would it possibly be 51k? It's a retarded value, you wouldn't use it in an application like this.
I've tried searching, but I haven't been able to find much. The left side of the diagram is otherwise relatively clear, 12V from ignition turns on BC547C, which in turn turns on BD140 from 12VBAT, which energizes the coil, but then what?
Maybe that coil is supposed to be the fuel level indicator, now that I think about it. The two middle lines go to the floater's cable from the fuel tank, but I don't understand the 00 resistor, or how the floater in the tank is coupled to the coil. Maybe the line from 00 is supposed to be taken as the data line for the indicator, it's just drawn wrong.

Anyone know if I'll have any issues using a solar power controller with a straight 12v supply? Or am I going to have to invert it first.

>>1735468
*12V DC
I really just want to use the controller for its ancillary functions like timing control

>>1735467
>don't understand the 00 resistor, or how the floater in the tank is coupled to the coil. Maybe the line from 00 is supposed to be taken as the data line for the indicator, it's just drawn wrong.
Yeah I don't get that either. My best guess is that there's a leading character that's gotten gobbled by the 0s. Though that doesn't really fit in with the other gobbled characters, typically the ones to the *right* are getting gobbled.

And yeah, I can't understand why a coil would be rated in ohms. Maybe that's supposed to be impedance?

What's the difference between these inductors?
Is FB10 a special type?

>>1735467
>but what inductors are rated as 1030 Ohm?
It's not an inductor, it's a relay.
And the other thing looks like a 33 ohm resistor being switched in and out, so if it's emulating fuel levels, it's only emulating two of them.
And why does it have two separate +12V power inputs?

>>1735478
FB = Ferrite Bead? I'm not used to seeing the inductance specified for ferrite beads.

>>1735478
FB suggests it's a ferrite bead inductor. And the dot indicates it's got "polarity" (from what I read inductor polarity isn't polarity in the sense of other passives, but that it has some sort of asymmetry that makes it behave differently when hooked up differently).

>>1735479
It's not marked properly if it's supposed to be a switch, I thought it was just a skewed line. I don't know why it has two power lines, I guess it's more "proper".
What is the relay doing? If that's not the fuel gauge, where does the line go to the fuel gauge?

>>1735482
>polarity
It might just be an indication that all the FBs should be wired in the same direction. It would be so they don't interfere with each other somehow, I guess. I mean, you probably would anyhow just from natural OCD, but maybe he expected a grunt from Roscosmos would install them backwards.

>>1735487
Well, a coil under a diagonal line between three points sure looks like a relay to me. What is it doing? Fuck if I know, I think the guy who designed it was the same guy who fucked it up in MS Paint. In other words, he wasn't any better at electronics as he was at graphics.
But it looks like it's switching the resistor (could be 33 ohms, could be 100) in and out.
>If that's not the fuel gauge, where does the line go to the fuel gauge?
I don't even see an input to the circuit. Also no kickback diode on the relay, I hope it doesn't kill the transistor after a while.

>>1735482
>>1735491
These are small SMD parts with no visible polarity markings. Does it matter if I put them the wrong in?

>>1735491
>Well, a coil under a diagonal line between three points sure looks like a relay to me.
Ah christ I didn't even identify that diagonal line as being a switch. I'm so used to seeing the contacts represented with solid dots I just assumed that was static wiring.

I'm thinking the guy who drew that schematic is dumb. I'd find something else.

>>1735499
Ferrite beads are almost always used for RFI suppression. I'm gonna say no. In my limited experience, the dot is to indicate polarity *relative to another coil* so that the AC phase will be correct. I'll bet the guy who put that schematic together didn't even notice the dot.

>>1734911
Good work Anon. I also CNC my own boards. Some tips: Eventually I started manually tracing the path I wanted the CNC machine to take because the automatic Gcode converter did a terrible job. It just outlines all traces so if you have two parallel lines with no ground between them it cuts the space between them twice. I found some free software called Gopt that optimizes Gcode so the machine isn't jumping all over the board while machining. Keeping things level is super important. I glue down a piece of wood and program the machine to level it with a flat head cutter. After that I've got a perfectly flat surface to put the copper clad boards on. Hot gluing them down from the sides while holding them down flat is a lot easier than clamps.

>>1735487
>two power lines
wild, senseless guess: some cuckmobile won't start if it thinks the tank is too empty and this is meant to fool it. one +12V is the always-on bus and one is the key-on bus
i agree it's a shitty schematic. OP should send it back for a refund

>>1735482
>the dot
it shouldn't matter unless you're doing something specific with the magnetic field

>>1735499
>3.9nH
thats not very much at all. make a little wire loop instead

Can having a buck converter dev board sitting on top a trace (no shorts) that carries signal for a digital led strip be disrupting that signal?

For some reason i can't figure out why the strip isn't working, i checked all traces and they are fine, there is power and everything going to both strip and the chip and all traces connect correctly, and when i pulled the chip out and tried it on a separate led strip it worked fine.

the data trace from the chip to the strip data pad is going under the buck

>>1735574
It doesn't seem like a great idea, but it wouldn't hurt to jumper the big bits of copper next to it to ground before starting over again.

>>1735574
If you already checked the signal trace isn't shorted to ground or VCC it could be possible EM interference from the buck is causing enough noise to disrupt the data signal, specially if the coil is just on top of the trace. You could try soldering a 10nF capacitor near the source of the noise or unsoldering the signal trace and run a wire around the board and see if that fixes it.

>>1735574
yes, that is called a transformer

>>1735579
>>1735581
>>1735590
welp, it must be it, since the buck has a coil on it the magnetism must be fucking with the signal since neopixels are super sensitive in regards to timing

And there simply isn't any other reason for it not to work, i will try to desolder the buck and connect it with wires which should give me my answer

>>1735600
Nah, just cut both ends of that trace and bypass it with a bodge wire to the side. A little hot glue to tack it down, and you've had a cheap and easy lesson on keeping power supplies separated from high-speed data lines. Seriously, this is a perfect example to learn from.

>>1735603
Oh yeah, you probably want to tack the wire over a (connected!) ground plane too. High speed data lines also do better when they're run along either a ground or a differential line.

TS100 or TS80? FIGHT!

>>1735625
T12 > TS80 > TS100

>>1735637
>handle is almost 20cm long
your number theory is confusing

>>1735639
>20cm long
Nothing wrong with that, could probably customise your own handle for the T12 if you're worried. Though the tip length itself is somewhat inflexible, so you could only drop about 6cm at the most. I'm considering getting one of those digital plug-in-mains irons with the buttons and screen on the handle for cases where I want cheaper tips or a second iron for tweezer/desoldering action.

My friend asked me to look at his HDD which isn't working. Found this which looks burnt out. What's the best way to try and fix it besides a new chip? Can I short between the dots?

hi /ohm/
got my hand on this old power supply block and want to use it to power a op-amp circuit (because of the +/- 12v output)
the thing was integrated in a custom made lab equipement, and it come with this 220v/220v transformer attach to it
anyone know what is it for ? can i just remove it and power the block directly with 220v ?
many thx !

>>1735683
The transformer looks like an isolation transformer, it improves safety by keeping the direct mains power away from the equipment. The power supply doesn't seem to have much amps to it, or the iso xfmr would be bigger.

>>1735680
Those two things look like zero-ohm links, so yes you theoretically could jumper across them. I'd still check with a meter to confirm they're open though.

>>1735680
From what I can see from here, those look like they're supposed to be fusable resistors, so if they are broken they might have done so because something else downstream is drawing too much current. I wouldn't replace them without checking out the rest of the circuit, especially not shorting them out.

>>1735693
0Ω resistors aren't supposed to heat up even a small amount under any kind of normal operations, so it's quite likely that they took the brunt of some kind of high-current event.
Also looks like they might have the number 3 on them, which could be a more definite resistance for fuse purposes.
They could also be there for current measurement, not sure how that would pan out but that via does make me a little suspicious.

>>1735683
Isolation transformer for safety. If they have one of those on there, I'd want to check the output rails (±12V and 0V) to see if they're referenced at all to either of the 220V inputs. If the output is properly isolated, I can't see why they'd use an isolation transformer. Then again, that implies a polarised plug of some variety, and neither the transformer's chassis nor the switcher itself have connections for a ground connection.

>>1732174
I never look at arbitrary circuits. Those sorts of schematics don't exist outside of textbooks. They're always labled in real life.

>>1735697
>Also looks like they might have the number 3 on them
Yeah you could be right. It's tough to tell without a clearer picture.

Honestly I'd see if I could source a HDD of the same model and board swap them if repairing that thing fails (or worst case, platter swap them and pray you don't have a head crash before you get all the data out).

(seriously don't platter swap if you can possibly help it)

>>1735704
I don't think it's that dodgy. Worst case scenario is tracing the circuit to figure out where those resistors are going through and looking up a datasheet to determine their purpose (and possibly even seeing them directly on the appnotes).

>>173569
Meter is showing 0 flow

>>1735697
I don't see anything else burnt out and can't find anything else with the meter, but it does look like the board suffered a "coffee event" once

>>1735704
Yea I found the chip on eBay for $70 but I'd rather not

>>1735709
Welp sounds like they're blown.

Take some isopropyl and clean the surface of them. Maybe some of that crusty shit will go away and let us see for sure if it's 3 or 0 (or something else).

>>1735697
You think those vias might just be test points?

>>1735710
How about the whole drive though? Unless it's fairly recent a used HDD should NOT run that much.

>>1735712
Nah he wants me to recover the data that's on it. It's a Seagate central with some fancy WiFi and network capability

>>1735711
Will try tomorrow, thanks!

>>1735713
>fancy seagate shit
the drive itself though? not the enclosure? i'm saying get the whole bare drive and swap boards. this is what recovery firms do.

>>1735641
length does kinda suck if you're trying to work under the magnifier
the very inexpensive T12 tips do appeal to my inner merchant tho, kinda thinking about retrofitting a 936 sort-of-clone with it

>>1735702
hear that, everyone? pussies don't reverse engineer

>>1735689
>>1735697
many thanks, continuity test show no conection between input and output pins, so i guess i will just remove the transformer
(it show some contuinity for the 200v input pins though, but i guess that expected)

>>1735752
Good to hear. Consider tying the 0V output rail to protective earth, if only for noise purposes.

Reasons to ground-fill everywhere you can:
>minimise relative capacitive coupling to all nodes but ground
>lower resistance to ground compared to standard traces
>shielding from external RF
>waveguide for internal RF
>pull heat away from components after soldering
>irk soldering wattlets
>consume less etchant
>consume less ink when printing transparencies

>>1735716
>i'm saying get the whole bare drive and swap boards. this is what recovery firms do.

not so fast, my negro. it's way more complicated than that. the PCB is matched to the platters at production time, so you need to remove the BIOS chip on the HDD and transfer it to the new PCB. other caveats apply, but this is the big one.

see here, for example, a HOW-TO for Seagate drives: https://www.hddzone.com/seagate_pcb_swap_replacement_guide.html

same site has diff how-to's for Maxtor, IBM, Samsung, and WD.

>>1735795
I bet those recovery companies have a few standard boards with breakouts for those wacky SMD sockets soldered onto them for the ROM chips to go on.

>>1735800

you didn't quite get the gist: those are EPROMS and data is generated and written to them just after production. this data is specific to the platters installed on that very drive, not some generic data.

>>1735805
I know, but desoldering and shoving in a socket is preferable to desoldering before resoldering.

>>1735805
Also don't you mean PROM?

>>1735509
Actually you don't need to do any of that manual leveling hogwash if you use probe leveling, which is what i am doing and it works great, plus the pcb itself isn't perfectly flat as well so if you are making your traces with a vbit then probe leveling is a must.
But if you are using a flat endmill then it doesn't matter that much.
>>1735603
>>1735606
okay i will give it a shot when i get home today, thanks
the only other thing i can think of would be the buck not having enough juice to handle the led strip since i stupidly didn't test it beforehand, but that shouldn't be it since the buck is rated for chink 2A, so that is european 1A, and the strip never draws more than .5A

>>1735816
>if you use probe leveling
Could you share what probe and software you're using? I think it's time I upgrade my machine. How does the probe deal with the end of the cutting head gradually wearing down and getting shorter?

>>1735848
>getting shorter
I assume it doesn't get shorter significantly from the start to the finish of a single engraving. You have to autolevel the coordinates for each board you carve AFAIK, so it wouldn't be much of an issue. Perhaps it depends on the material of the cutter though, not sure if carbide or HSS or whatever are chosen.

>>1735725
Reverse engineer? What are you, Chinese?

>>1735854
how else will you be able to tell whether chinks stole your design, cuckwad

>>1735848
That is not how it works.
You attach a signal wire to the endmill and a ground wire to the pcb and then it moves around the pcb and stabs it in a grid pattern so it knows that for example left top corner of the pcb is 1 mm higher than the top bottom corner and so when you are milling later on and the gcode says 'mill into left top at -1mm Z axis' the mill will actually execute the milling at 0mm Z axis since that corner is higher

>>1735812
implementation detail. they are probably erasable, they just never are in practice

>>1735800
or, hypothetically, the matching really isn't all that important if you don't write to the disk

>>1735702
>never opened a random equipement repair manual ever
Sure, the A1 sized page has a "Power supply circuit label " on top but that is it usually.

>>1735859
>they are probably erasable
Yeah but that would mean an EEPROM, not an EPROM. EPROMs are the edgy ceramic ICs with quartz windows.

>>1735795
Yes this is true. There is matching involved. Still less risky than platter swapping though.

Can you make a smt component counter that counts how many components you're pulling from a strip by counting the lateral holes and multiplying by 1, or 1/2, 1/4, or 1/(number of side holes for each component)?

>>1735929
>lateral holes
you mean the sprocket holes? probably, if the pocket positioning is pretty uniform. they do this with photographic film in late-model film SLRs: there's an IR LED and photoresistor pair and the camera counts the number of pulses. alternatively you could use a wiper and a contact and count the pulses (this is how punchcards are read).

I wouldn't be surprised if pick and place machines advance the tape using such a method, even if they use a more refined method to align with the individual part.

>>1735929
Honestly that is what I'd do if I was tasked with that. Should be easy enough to do with a LED and a receiver pair.

>>1735934
Yes, the sprocket holes. There are stips that are clear, so I don't think I can read the strip with the IR led directly.

Wiper and contact sounds good, since I just happen to have steel dowel pins and pogopins.

>>1735941
The problem is that we also use smt components with clear strips, so a LED will probably not work

>>1735600
>>1735574
>>1735603
Fuck.. so i cut the data trace on both ends and replaced it with a wire and also connected the ground to the outside copper are and it's still not working... this is driving me crazy

>>1735943
Uh, I was thinking of a infra red LED (as I usually only have photodiodes for that range). Maybe whatever plastic you are using is opaque to IR

>>1735947
And even if does not block 100% of the IR you'll probably notice a dip when a hole passes and you just need to amplify that and compare it.

>>1735947
Only one way to find out

Aaaand

The black strips work ok with the IR sensor

The clear strips don't

So I've gotta use the pogopin+dowel approach

>>1735955
Also, the paper strips won't work with the pogopins, so I need to make a hybrid counter, one side with an IR sensor, one side with pogopin/dowell

>>1735959

an optical mouse detector would work, but dunno how accurately.

>>1735959
>the paper strips won't work with the pogopins
The paper conduct too much or that it tears/gets caught on the paper? If it's the latter, you might try literally just using thin wire/conductive fiber (like steel wool). You don't need much of a contact to generate the pulse. And instead of being pressed into the strips by a spring, the brush is more gently dragged out of the way by the passing tape.

>>1735955
Are you sure there isn't even a noticeable change in transparency? Remember that amplifying things a couple thousand times is easier than finding a new transducer

>>1735945
okay, so, i oscoped it and shit just got downright riddiculous

bottom pic is signal from the chip (fartduino) being in the breadboard where i tested it on my led strip spool and it works fine there, the leds light up, so i know that is how the signal is supposed to look

top pic is the signal from the PCB in the photo.
As you can see they are identical, so the signal is 100% correct and going into the led strip data pad, so why the FUCK isn't it working. I even oscoped the buck converter output to make sure there isn't any voltage drop or noise, and nope, the output takes less than 1ms to jump to 5V and then it stays there nice and clean, as expected.

So the strip is getting 5V, GND, and correct DATA sequence and yet... it's not fucking working.
I even probed the beginning and end of all 3 pads to make sure the traces in the strip aren't broken

pls help, i am about to drink my entire bottle of fecl3 because i can't take this anymore

>>1736001
okay...
...
...i just fixed it...
..
but i am not going to tell you how because it's so fucking embarrassing i would legit have to kill myself .
Thank god this board is anonymous.

>>1735945
>>1736003
This shit is honestly why I went with software over hardware. It took me many years to develop the discipline to deal with hardware, going step by step and constantly testing to avoid it flaking out in random obscure ways.
WIth software it's easy to undo undo undo on your source and see when the thing died.

>>1736003
Kek. I've have one of these problems once a month. I think most hobbyst/engineers do.

>>1736016
Oh i fucking hate electronics debugging.
In software you just slap in some print statements or break points and done
In hardware you have to solder in fucking wires and connect oscope probes and multimeter shit and, desolder shit, etc etc, god it's so frustrating sometimes.
And of course there is no danger of your program getting fucked up by random noise from a damn refigerator nearby

>>1736020
Yeah, well actually I ended up as an embedded systems programmer, because I knew enough about electronics. One of my first tasks was to talk to a liquid tank level sensor thingy with a one-byte protocol, from our 8-bit microprocessor system. This was 1997, and in assembly language. I had a comm scope and just dumped extra bytes onto the line because who the fuck cares if I confuse that thing when I can reach over and reboot everything.

>>1736031
Please tell me you used a 6502

>>1736032
Fuck no it was a fucking 6809, god king emperor of the 8-bits.

>>1736033
>shitarola
F

>>1736038
6502 is a fucking aspie chip, and a quarter of the opcode space would lock it up
6809 was designed to run position-independent C code, and it only locked up if it saw a $14 or $15 opcode (and that's how I found out that our watchdog circuit was being reset by reads as well as writes)

>>1736045
>a quarter of the opcode space would lock it up
well to be fair everyone with a brain uses the CMOS version, which just treats all the illegal opcodes as NOPs

>>1736031
My favorite thing in electronics is always writing the code for the microprocessor. Wiring the thing up afterwards and soldering all those resistors and other such not so much, but can't get around that.
So basically i love the digital, hate the analog.
If we were still in the analog age i am pretty sure i would never even get into EE

>>1736064
Yeah, CompE was always more my forte. But I think my main problem was I just could never really visualize what the fuck the electricity is doing in a circuit. Like I never really got past the water pressure analogy.

I think not taking physics in high school or college might've been where I went wrong.

>>1736070
I wonder if plumbers get confused when they teach them how water flow and pipe pressure works by using the electricity analogy

>>1736070
>>1736064
I'm basically the oposite. I like soldering and building stuff but every time I have to program a micro I just cringe. It's just me being annoying because I get motivated after a while when I see things taking shape. I didn't write a AD - > 7-segment display voltage meter for my bench supply out of lazyness tho.

>Page 5
Wew, finally found it.

Does tip tinner like this do shit? My brother got me a PlaySkool soldering station for Xmas, you know that blue and yellow one, and I would like to not destroy the tips the first time I heat them up.

I’m stepping up from a $4 Chicago Electric with no temp control, but I don’t understand the whole tip tinning thing because the solder just balls up at the end when the iron is heated up. Where have I gone wrong?

Also the tips, there’s a million 5pks of tips for like $10 from random brands, or the Japan imports for $25, is there a difference? I’ll gladly buy the $25 tips if I can learn how to make them last.

>>1736111
smoking is bad

>>1736111
>tip tinner
>plot twist : it's just solder

>>1736045
>crashes when you use the cheap but not-too-cheap opcodes
Motorola in one post

>>1736111
yes, it does great shit when your tip is so oxidized it can't pick up solder
>just balls up at the end
you don't need tin to cover the whole tip. just the faces you want to use. in fact there are some bevel tips that are tinned only on the face, meant for drag techniques. on those, solder doesn't stick to the sides of the tip and possibly get all over everything

>>1736124
>second plot twist: and some extra-strong oxidizer

>>1736124
It kept popping up on the “frequently purchased together” lists, it was $6, and next day delivery.

I would rather go through trial and error and shitpost about it than read proper directions, but I did get a good tin on the cheap tip the one time when I caught it as it was cooling off. But the next time I got the thing heated up, soldered a couple wires, and then left it sitting turned on for a couple minutes, it got black and crusty and oxidized again.

I thought I read something like the oxidation happens with heating and cooling, is that true? So it doesn’t get rekt while it’s turned on, but rather as it heats with no tinner on the tipper?

>>1736125
Ahhh so it’s balling up because the tip is already oxidized?

>>1736125
>opcodes
It's the "test" opcodes that Motorola for some reason back in the day decided were a good idea. The CPU just goes into a loop reading every memory address. No flag to set first, no page opcode, no write 55 or AA to a register, nothing. The only way out is with RESET.
Basically it was the HCF trope made real.
The fun part was using an ICE that shared a single CPU with the target.

>>1736125
Sooo how exactly do I use this? Didn’t come with any Ikea-esque instructions. Do I roll it around as it’s heating up? Right before I wanna solder something? Or at the end when I’m turning it off?

From an LED light strip, started flickering and this bitch was bulging. Too lazy to get a new cap but wanted to see if I could fix it.

How close would I have to get for the values? Will anything else work if I can stay close?

>>1736184
The rule of thumb with caps is that if you can't get the same values, you can go up, but you shouldn't go down.

>>1736197
I think I snagged a 100V 100uF out of something and was going to try it just to see, but I gotta be up at 4am and don’t want to open that thing up again, I hate going to bed with half a project on the workbench.

>>1736201
>replacing 330 uF with 100 uF
I wouldn't.

>>1736197
>shouldn't go down
too late, tripfags can't stop sucking cock

>>1736203
What about the voltage rating? Capacitance is more important, but go above?

It’s junk anyway, I was only hanging onto it because I thought maybe I could play with the bright LED strip, but it wants quite a high voltage I believe.

>>1736255
It’s good side money

>>1736257
Yes, the rule of thumb is that both values can go higher. I'm sure there are reasons why and situations where you should match values, but my understanding is that in typical applications where you see electrolytics you can go with higher-than-rated values.

>>1736111
>>Page 5 Wew, finally found it.
protip newfag, bookmark this: https://4chan.org/diy/ohm